Characterization or analysis of samples (e.g., thickness of a thin film, elemental and/or chemical species concentration in a thin film formed on a substrate, etc.) is necessary in the manufacture of many different types of devices (e.g., electronic and optical electronic devices). For example, it may be necessary to determine the composition of thin dielectric films (e.g., gate oxide films, tantalum nitride films, etc.) formed in known semiconductor integrated circuit devices, such as processing devices and memory devices. Increases in the density of such devices on an integrated circuit chip and reduction in device dimensions require the advancement of production processes and characterization technologies related to the materials used to fabricate such devices.
For example, recent developments in the fabrication of semiconductor devices may employ shallow implant and/or other ultra-thin structures. In one particular example, gate oxide layers have become very thin films, typically in the range of about 1 to 10 nanometers in thickness. Such thin films are difficult to characterize. Such structures will require characterization techniques that have improved sensitivity over conventional characterization techniques.
Further, such techniques may also require the characterization to be performed with ample speed. For example, when such characterization techniques are used to monitor manufacturing tools or processes, e.g., metrology for wafer level manufacture of various films on substrates, the characterization must be done at suitable process speeds. Further, in monitoring of the manufacturing process, preferably, it is desired that such characterization be performed in a nondestructive manner, e.g., using a noninvasive process.
In addition to performing such process monitoring characterization of thin films noninvasively and with ample speed, one must be able to carry out such characterization techniques with precision on a consistent basis. In other words, measurements made or parameters determined using the characterization process must be repeatable with a suitable measurement precision (e.g., relative standard deviation, RSD). In such a manner, a process excursion, e.g., a process not performing as it is intended which typically results in product being produced that is not acceptable, can be detected.
Various techniques have been used for characterization of materials, e.g., to provide thickness measurements and/or to determine the concentration of trace and/or major components in such materials. For example, several of such methods include ellipsometry methods, transmission electron microscopy (TEM), scanning transmission electron microscopy (STEM), secondary ion mass spectrometry (SIMS), x-ray photoelectron spectrometry (XPS) (also known as electron spectroscopy for chemical analysis (ESCA)), Auger electron spectrometry (AES), and other electron beam methods.
Many of such techniques are sensitive to the near-surface region of a material. Further, many of these techniques also permit a measurement of material properties as a function of depth beneath the surface through depth profiling. In typical depth profiling, for example, continuous or periodic ion beam sputtering removes material from the surface of a sample to expose progressively deeper material at one or more various depths of the sample for further measurement and/or analysis. Generally known sputter rates may be used to determine the depth at which the surface measurements are completed. As such, a characterization of the sample as a function of depth beneath the surface can be attained. However, in process monitoring, such depth profiling techniques are in many circumstances inadequate. For example, depth profiling is an invasive and destructive process, and further, such processes generally take a relatively long period of time to complete, e.g., as compared to just surface measurements.
Further, even if such techniques are not used in a destructive depth profiling fashion, they are typically inadequate in many other respects with regard to characterization of various types of samples. This is particularly the case with respect to characterization of thin films, e.g., thin gate oxide films.
Optically based ellipsometry methods have been the standard method for monitoring SiO2 gate films. However, the change in gate materials, from for example, SiO2 to SiOxNy, has made a major impact on the usefulness of optical measurement tools. The index of refraction changes when nitrogen is added to the film. Since nitrogen content varies with depth, the index of refraction of the films is a variable making it difficult to use standard optical methods to monitor SiON films. In addition, the trend to thinner films (e.g., 18 angstroms, 12 angstroms, 8 angstroms . . . ) is challenging the fundamental limits of optical methods. The combination of these two effects has reduced measurement precision for thickness using such techniques rendering them ineffective for monitoring thickness and composition. Currently, for example, optical techniques achieve precision of 0.4% RSD for thin (e.g., less than 20 angstroms) un-nitrided oxide (e.g., SiO2) films and 1.5-10% RSD for nitrided (e.g., SiON) oxide films.
TEM or STEM combined with electron energy loss spectroscopy (EELS) measurements can also provide thickness and some composition information. However, there are a number of issues that make the TEM impractical for use in production monitoring. For example, thickness measurement precision is typically greater than 2 angstroms and the cost of the needed equipment is generally prohibitive. Further, the length of time needed to perform such measurements is long (e.g., four hours per measurement) and a highly skilled specialist to prepare the sample and perform the measurements is typically a requirement.
Further, for example, SIMS, which has a very small sampling depth, is routinely used to quantify low level dopants and impurities in thin films (e.g., thin films less than 10 nanometers) because of this technique's extreme surface sensitivity (e.g., single atom layer sensitivity and ppm-ppb detection limits). However, sensitivity factors used for SIMS quantification are matrix dependent and accurate quantification requires the use of calibrated reference samples. For example, when the concentration of a dopant exceeds 1%, it becomes a significant part of the matrix further complicating the task of quantification. To monitor silicon oxynitride gate films via SIMS, it would be required to regularly (e.g., at least daily) analyze reference silicon oxynitride films with thickness and nitrogen dose certified by an external direct measurement technique such as XPS.
Further, AES has also been used for thin film characterization. However, the high intensity electron beam used to make Auger measurements can alter the apparent composition of a thin film by causing chemical damage (e.g., can damage SiO2 films) or causing the migration of elements within the thin film. For example, there are concerns over the possible mobility of nitrogen within a film under the influence of an electron beam (e.g., nitrogen is known to migrate to the interface of an oxide-nitride stack (ONO) provided on silicon).
XPS, or ESCA, has been previously used to characterize thin films (e.g., ultra thin films less than 5 nanometers) such as lubricant coatings on computer hard disks with a measurement precision of 5% RSD. Further, characterization of other types of films such as SiON via XPS using standard practices has resulted in measurement precisions of 0.5% to 1.0%. For example, such standard practices involve the collection of data at relatively low analyzer angles such that depth resolution is enhanced. Such a low analyzer angle is typically less than 20 degrees. Use of a low analyzer angle generally results in a slow characterization process and also may result in problems associated with placement of the analyzer of the characterization system relative to the sample being analyzed. Yet further, existing data reduction methods which operate on the XPS collected data employ software tools for background subtraction and peak fitting that require frequent operator input making the results operator dependent and less precise.
In general, many of the techniques described above for characterizing thin films are invasive techniques, e.g., they involve destruction of at least one or more portions of the sample. Such techniques, e.g., those that use removal of material during depth profiling, are sufficient in many circumstances, e.g., research and development, product testing, etc., but do not provide for the ability to quickly analyze a thin film such as is necessary in production processes. For example, in such production processes, a thin film being formed typically needs to be analyzed so that such information can be used for production control, product test, etc., without loss of product due to invasive characterization of such films.
Nitrogen doped or nitrided silicon oxide is one material that is used as a gate oxide for a transistor structure. Such gate structures are only one of the growing number of semiconductor related material structures under development that require characterization at an unprecedented level of complexity. Such challenges are not limited to merely a desire for near-atomic and monolayer spatial resolution, but are magnified by the level of accuracy, precision and speed demanded by the semiconductor fabrication industry.
There is a distinct need to develop adequate characterization methods and systems. The ability to characterize materials at such levels is necessary to enable product development and also necessarily precedes evolution of process control. For example, there is a need for suitable systems and methods to provide parametric thickness, nitrogen dose, and nitrogen distribution information for thin nitrided silicon oxide films such as used for transistor gate oxides.